Marketing Research is About Testing, Testing and More Testing

A common theme that runs throughout the material found on KnowThis.com is the need for marketers to engage continually in research. We especially mentioned the importance of research last June when we reported on how marketers are using a special method of research, called A/B testing, to test online marketing efforts, such as advertising, pricing and webpage design. In that post, we noted that A/B testing is a form of experimentation where a marketing decision is studied to see what works best. For instance, a test may involve determining where on a website an advertisement will generate the highest revenue. While technically A/B testing only involves testing two different options to see which is better (e.g., ad located in upper left vs. ad located in top middle), testing involving the manipulation of more options (called multivariate testing) is also widely used.

In fact, A/B-style testing has become one of the most dominant forms of online marketing research. This is because it is relatively inexpensive, can be created very easily, and provides useful information in a short amount of time. A/B-style testing has grown in part because of online research facilitators such as Optimizely, which enables websites to test different layouts to see which one is more effective. Google is also a major provider of A/B-style testing. Advertisers using Google’s AdWords service can carry out tests with the Content Experiments feature while websites that generate money by placing Google ads on their site can run experiments to test different ad designs.

However, A/B-style testing is certainly not limited to the online world. It can be used in many other offline situations including testing product layout in stores and testing different types of ads on television. An interesting example of how A/B-style testing is done with television ads can be seen in this CNN Money story. The story discusses how a lingerie retailer, Adore Me, tested television ads to see how model’s hair color (blonde vs. brunette) and body dimension (standard-fashion-model vs. plus-size model) affected lingerie sales. The tests revealed that an ad featuring models with blonde hair and standard-fashion-model dimensions was much less effective in generating sales compared to an ad with models with brunette hair and standard fashion-model dimensions. They also found that an ad featuring models with blonde hair and standard fashion-model dimensions was less effective than an ad featuring a model with brunette hair and plus-size dimensions.

Certainly this information will help guide Adore Me’s television advertisements. However, it is also important to understand that the information obtained should not suggest that this is the way the market will always view this issue. That is why marketers cannot rest on a single set of research results and, as the founder of Adore Me acknowledges, must continually carry out research because the market is always evolving.

Image by mil8

Some Thought This Would Never Happen but Older People Are Starting to Embrace Social Media
How Online Sellers Are Using Experimentation Research